Background
Cornelia de Lange syndrome (CdLS) is caused by mutations of the NIP-BL gene on chromosome 5p13.1 for nearly 60% of individuals and by mutations on theSMC3, RAD21 and the X linked SMC1A and HDAC8 genes in a smaller proportion of affected individuals [
1‐
5]. All five genes are involved in the structure and regulation of the cohesin complex which is crucial for neural maintenance and repair [
2,
6]. It is probable that there are further unidentified mutations relevant to the cause of CdLS [
7]. The physical phenotype of CdLS includes low birth weight, small stature, limb abnormalities, distinctive facial features, hearing and vision abnormalities, and cardiac, genito-urinary and gastro-intestinal disorders [
8‐
10]. Degree of intellectual disability is variable and typically severe (30%) to profound (45%) with poor expressive relative to receptive language [
11‐
14].
Behavioural research has focused on self-injurious behaviour and, more recently, autism spectrum disorder (ASD) characteristics (e.g. [
13‐
16]). Reported prevalence rates of autism spectrum disorder in CdLS range from 43 to 67% [
13,
14,
17‐
21]. However, the presentation of the triad of impairments in CdLS may not be typical of that observed in idiopathic ASD [
19,
20]. Specifically, social impairment in CdLS is characterised by selective mutism, extreme shyness, social anxiety and social avoidance [
11,
19,
22‐
24]. Although repetitive behaviours do not appear to contribute as significantly to the ASD profile, relative to social interaction and communication impairments [
25,
26], they are characteristic of the syndrome, specifically lining up and tidying up behaviours and a strong preference for routine and need for sameness [
15].
There is emerging evidence of age-related changes in behaviour in CdLS. Kline et al. [
27,
28] reported 80% of individuals to show high levels of depression, self-injury, obsessive-compulsive behaviours, anxiety, aggression and hyperactivity. Blagowidow, Kline and Audette [
29] also reported that behaviour disorder increased at puberty. Basile et al. [
17] reported increased behavioural changes with age including communication disturbances and anxiety in 56 individuals aged 11 to 31 years, and Sarimski [
12] found children over age six experienced significantly more social isolation and anxiety than younger children. Oliver et al. [
25] reported that adults with CdLS were more likely to experience levels of negative affect and impulsivity (13%) compared to children with CdLS (3%), a profile not reported in other syndrome groups. Low mood is also reported within the teenage years compared to fragile X and Cri du Chat syndrome groups and is associated with higher scores on measures of insistence on sameness [
30]. In combination, these observations of the behavioural phenotype of CdLS and its correlates suggest cognitive difference and change with age might be evident in this syndrome and related to the behavioural presentation.
Comparatively, less research has been published regarding the cognitive profile of CdLS. In this study, we will describe the cognitive profile, specifically executive function, in more able people with CdLS and consider the relationship with chronological age and its association with behavioural change. Executive functioning refers to cognitive abilities that control and regulate other abilities and behaviours and are necessary for goal-directed behaviour. They include the ability to initiate and stop actions, to monitor and change behaviour, and plan future behaviour when faced with novel tasks and situations. Executive functions allow individuals to anticipate outcomes and adapt to changing situations. The ability to form concepts and think abstractly are often considered components of executive function [
31].
Specific deficits in executive functioning are thought to account for some of the observed behavioural problems seen in autism spectrum disorder [
32] and are also described in a range of genetic syndromes such as fragile X (e.g. [
32‐
36]), Prader-Willi, and Down syndromes. Woodcock, Oliver, and Humphreys [
37] reported that impairment in attention switching in Prader-Willi syndrome was related to behavioural reports of adherence to routine and temper outbursts, providing evidence for an executive functioning-behaviour link. Similarly, changes in executive function that accompany decline in Down syndrome have also been described [
38].
The frontal lobes are postulated to play a major role in executive functioning [
39] and are the last part of the brain to fully develop. Studies of Prader-Willi, fragile X, and Williams syndromes reveal abnormalities in the frontal region that may be related to compromised executive functioning [
35,
40]. For example, fMRI scanning of individuals with fragile X, in which difficulties with inhibition and visual attention switching are evident [
35,
41], has shown reduced activation in the prefrontal cortex [
42]. Brain imaging studies of individuals with CdLS are lacking. However, the few available autopsy studies have revealed frontal lobe hypoplasia in CdLS, indicating possible disorders of axonal growth, neural priming, and neuron cell repair [
43]. Therefore, it may be postulated that impaired growth of the frontal lobes or emergent neuropathology may underpin behavioural change seen in CdLS, especially that occurring during adolescence and early adulthood [
44].
Recent research by Gimigliano et al. [
45] analysing the proteomic profile of the SMC1A and SMC3 genes implicated in CdLS showed protein expression was dysregulated. CdLS cell lines were found to show an increase in global oxidative stress as a result of this dysregulation, which the authors postulate could contribute towards the phenotype of the disorder, including premature aging and associated cognitive changes. They discuss how the subsequent reduced antioxidant defences could ultimately lead to cell death through DNA damage, membrane potential loss and reduced synthesis of ATP. These conclusions are tentative and require further investigation; however, the role of oxidative stress in neurodegeneration has been evidenced in other disorders. For example, oxidative stress in the brain has been associated with the pathogenesis of neuron degeneration and death in Alzheimer’s disease [
46,
47]. It is therefore possible that oxidative stress may cause a chain of reactions that could account for the neuropathological and cognitive changes occurring with age in CdLS.
In summary, the available evidence suggests that behaviours that are phenotypic of a genetic syndrome may be underpinned by executive functioning impairments. The behavioural phenotype of CdLS also suggests a description of executive function deficits would be informative. Furthermore, reports of behavioural changes with age in CdLS, specifically around the age of adolescence/early adulthood, suggests that examining the association between executive function and age in CdLS would be informative as well.
A contrast group of individuals with Down syndrome (DS), comparable for age, gender, mobility, level of adaptive behaviour and receptive language (a domain not thought to tap into executive functioning [
48]), was included in the current study. Down syndrome has a prevalence of 1 in 600 live births. The syndrome is caused by an extra 21st chromosome in 95% of people affected [
49]. The physical phenotype includes epicanthic folds, protruding tongue, flat nasal bridge, brachcephaly, broad hands, brachydactyly and lax ligaments [
50]. Developmental delay is also prevalent. At age 21, mean IQ is 42 (range 8–67) [
50]. According to other studies (e.g. [
51]), the behavioural and cognitive phenotype for DS includes relative strengths in elements of visuospatial processing [
52,
53] and social functioning [
54,
55], alongside relative deficits in language [
56], verbal processing [
57], verbal short-term memory and explicit long-term memory. However, visuospatial short-term memory, associative learning and implicit long-term memory functions are relatively preserved [
58]. Executive functioning skills are considered to be broadly impaired adults with DS. Specifically, deficits in inhibitory control [
59], flexibility [
60] set shifting, sustained attention and planning have been documented in adults [
61], alongside the aforementioned deficits in short-term memory. Fewer studies have evaluated these skills in children, although studies indicate that the profile of impairment may be similar to that observed in adults [
62,
63]. According to Pritchard et al. [
64], deficits in executive functioning in children and adolescents with DS are largely mediated by associated co-morbidities such as autism spectrum disorder symptomatology and ‘disruptive behavior disorder’. The role of intellectual disability in this association is not fully understood, and findings are inconsistent [
62,
64].
In this study, a well-matched, homogenous group of people with DS, rather than a heterogeneous group of people with intellectual disability, was considered an appropriate contrast group. The literature describing cognition in individuals with DS is fairly consistent, and more is known about the behavioural and cognitive phenotype of DS than any other syndrome group. Therefore, this population provides a useful benchmark for profiling and interpreting the strengths and weaknesses in CdLS. This group can be positioned relative to the known areas of difficulty in individuals with DS. Due to the relatively small sample sizes, such a homogenous group was considered to add greater statistical power.
Discussion
This study details the results of the first evaluation of the executive functioning profile in more able adolescents and adults with CdLS. Participants with CdLS were compared to participants with DS, using several measures of executive functioning. Initial comparisons indicated that the two participant groups did not differ with regard to gender, chronological age, adaptive behaviour and receptive language age equivalence scores.
Carer-rated BRIEF-P
T scores revealed that participants with CdLS demonstrated a similar profile of executive functioning to the DS group but with significantly greater impairments in flexibility. This suggests that individuals with CdLS are broadly compromised in executive functioning skills, to a similar degree as has been previously demonstrated in those with DS [
78]. On some subscales, scores in the CdLS group approached clinical significance (e.g. Shift and Working Memory). As raw scores were normed on a group of typically developing children who were younger than reported age equivalence scores of some of the CdLS and DS participants, the magnitude of executive functioning deficits may have been weakened. This could explain why clinical significance was not reached. Currently, there is no informant report assessment of executive functioning which considers both a person’s chronological age and their developmental age, which is something future studies should address. The fact that individuals with CdLS are significantly more impaired than the matched DS sample in the Flexibility Index is indicative of a specific syndrome group difference. Direct assessment, using the DCCS, confirmed that participants with CdLS demonstrated significantly greater difficulty in flexible thinking compared to the DS group, continuing to sort cards by the first rule of the task, rather than shifting to the new rules as the assessment progressed. Given that there were no significant group differences in phonological or spatial working memory (as measured by the backwards Digit Span and Corsi Span tests, respectively), the difficulties in rule shifting evident in the CdLS group are unlikely to be related to learning and retaining the rule of the task but in perseveration and inflexibility. Equally, it is unlikely that these differences are attributable to degree of disability, as both groups were well matched on receptive language, adaptive behaviour skills and developmental quotient. In relation to working memory skills, both groups were limited in their span capacity expected given their chronological age (both having spans of three). In typically developing populations, working memory capacity typically increase from a span of three at 4 to 5 years of age to a span of seven to eight at 16 years [
79]. Impairments in capacity of the phonological loop impact on speech and language development [
80], which is consistent with the developmental profile of these two syndrome groups. Further analyses suggested that both groups were more impaired on verbal working memory relative to spatial working memory; however, without a typically developing sample, it is difficult to determine whether this is evidence of atypical working memory abilities in these syndromes. Individuals with CdLS were significantly more impaired than individuals with DS on the forward span tasks (across both domains) while both groups were similarly impaired (and in fact scoring almost at floor level) on the backward span tasks. For the DS group, the increase in the number of individuals scoring at floor from the forwards to the backwards span tasks further demonstrates difficulties in working memory across both domains. A similar effect was observed in the CdLS group; however, there were notably more individuals scoring at floor level on the forward span tasks than the DS group. This may suggest individuals had difficulties understanding task demands; however, the decrease in scores when switching from the forwards to the backwards tasks is consistent with the increase in difficulty which suggests the low scores are an accurate reflection of the individuals’ difficulties in task performance. Verbal fluency was marginally poorer in the CdLS group compared to the DS group, while both groups had difficulty with the more complex fluency tasks which required a greater capacity for organising concepts in a novel way [
34]. The differences in verbal fluency are particularly interesting given that there were no significant group differences on the receptive language skills that might confound performance on such tasks. Results from the design fluency task revealed little difference between the two groups. Both groups appeared to show difficulties in monitoring their drawings, with many repeated designs being drawn. This suggests that difficulties in planning and monitoring are characteristic of both syndrome groups and this might be accounted for by global cognitive impairments rather than a syndrome-related executive function deficit. However, further investigation with a matched typically developing contrast group would be needed to fully determine the nature of this deficit in both groups.
Overall, there is large overlap in the executive functioning profiles of individuals with CdLS and DS. However, there are also marked group differences on very specific domains of executive functioning which may represent syndrome specific deficits in CdLS. The profile of differences in executive functioning in individuals with CdLS is consistent with higher prevalence of repetitive behaviours, specifically tidying up and lining up behaviours in individuals with CdLS and the strong insistence on routine and need for sameness described by Moss et al. [
15]. According to Turner [
81], these behaviours may all be explained by deficits in executive functioning skills. This theory is supported by findings that deficits in executive function are observed in a range of other neurodevelopmental disorders with increased repetitive behaviour including autism spectrum disorder, Prader-Willi syndrome and Fragile X syndrome [
32,
37,
82].
Some of the CdLS participants failed to engage in all of the tasks. Those who did not complete the tasks were not significantly different to those that did, with regards to the characteristics described in Table
1 (
p > .05 for all variables). Since there were no overall group differences in verbal working memory, it is unlikely that failure to complete tasks was due to weaknesses in recalling instructions. It is tentatively suggested that problems with task initiation may account for the failure of some participants to engage in the tasks. This is supported by the authors’ clinical observations and by anecdotal research observations describing difficulties in task initiation [
83]. It needs to be determined whether this is due to social/performance anxiety that is also characteristics of this group [
84] or whether this is determined by the specific executive functioning difficulties that appear to be characteristics of this group. However, it is not clear why these difficulties in task engagement were not consistent across all executive functioning tasks.
A significant correlation between age and backwards digit span suggests that older individuals with CdLS have poorer verbal working memory compared to younger individuals with CdLS; although, caution should be taken when interpreting this finding given the exploratory nature of the analysis. This pattern of results is consistent with previous research which has also highlighted changes in mood, behaviour and cognition with age [
26,
83,
85] and reported physical signs of premature ageing [
28]. The compromised function of the cohesin pathway (resultant from the genetic mutations which cause CdLS) has been implicated in these behavioural and cognitive changes in CdLS, due to the role of this pathway in neural maintenance and repair [
28]. Recent evidence also indicates downregulation of proteins involved in the response to oxidative stress and an increase in global oxidative stress in CdLS cell lines which may be directly linked to the phenotypic changes in the syndrome [
86]. A larger sample with a wider age range and prospective longitudinal follow-up is needed to examine the association between age and executive functioning in more depth as it may be that individuals in the sample were already experiencing some age-related changes so correlations may have been weaker as a result. It is of course also worth considering ascertainment effects in relation to age-related findings. Greater understanding about the condition and better diagnostic testing may lead to earlier diagnosis, and identification of milder forms of CdLS than was previously the case, so it may be argued that any differences related to age could be attributed to the older groups having more severe difficulties. More research with larger samples is clearly indicated.
The findings should be considered within the context of some methodological shortcomings. Primarily, the use of a group of participants with DS warrants comment. One difficulty with using this group is their increased risk of Alzheimer-related dementia. Alzheimer-related dementia in DS has a prevalence of 0–2% in individuals under 40 years old and more than 40% in those over 60 years [
58]. The later stages of dementia in DS are well documented in the literature while research into the earlier and likely more subtle cognitive and behavioural changes that may occur in the initial stages of dementia in DS has only started to emerge in recent years (see [
38]). Although the DS sample in the current study were all under the age of 40, thus mitigating the likelihood of dementia, it remains a possibility that some of these individuals may have been affected by the early stages of dementia which could impact on their performance. Independently of the association with Alzheimer-related dementia, research studies suggest that both children and adults demonstrate broad impairments in executive functioning skills [
60‐
63,
87]. Therefore, where no significant group differences were identified in the current study, we can determine that individuals with CdLS demonstrate a similar degree of impairment in these skills relative to those with DS; however, we cannot determine fully the extent of this impairment and how it is related to the presence of an intellectual disability. Future research should include a typically developing contrast group that would enable the degree of impairment in CdLS to be determined more precisely. However, with both of these limitations in mind, it is poignant that the CdLS group performed significantly worse than the DS group on specific aspects of executive function and noteworthy that the two groups demonstrated equal abilities in these areas.
The sample size of each group was relatively small which may have comprised statistical power. Also, whilst a variety of tests were used to examine different aspects of executive functioning, there are some elements of executive functioning that have not been addressed in the current study and as such will need to be looked at in future research to complete the picture of executive functioning in these groups. There was also no measure of motor abilities in the current study. This would be a useful measure to use in future research so as to rule out difficulties in motor skills as explaining difficulties with tasks requiring the participants to draw or write, for example Design Fluency.
Acknowledgements
We are extremely grateful to all of the participants and their families who gave their time to take part in this research study. Thanks to Leah Bull and Sarah Gorniak for their help with the data collection.